Re: [patch] Real-Time Preemption, -RT-2.6.10-rc2-mm3-V0.7.31-19

[Florian Schmidt <mista.tapas@gmx.net>]

On Thu, 2 Dec 2004 09:40:40 +0100
Ingo Molnar <mingo@elte.hu> wrote:

> > actually a single client doing nasty (non RT) stuff in its process()
> > callback can "starve" jackd. AFAIK jackd waits until the last client
> > has finished its process callback. So, if some client's process
> > callback decides to use (for example) some blocking system call (big
> > no no) and consequently falls asleep for a relatively long time, then
> > it can cause jackd to miss its deadline. I'm not sure though wether
> > this triggers an xrun in jackd or just a delay exceeded message.
> 
> since the period size in Rui's test is so small (-p64 is 1.6 msecs?),

ca. 1.3ms

> the 2.6 msec timeout in pipe_poll() already generates an xrun.

yep. now the question is: why did jackd have to wait so long for the
client? what was the client doing? was it sleeping? what client was it?
probably not the simple jack_test client, right?

Short tests ((20 minutes, due to lack of time) with cvs jackd and a
number of modified jack_test clients (added a for loop in the process()
callback to burn cycles to increase the cpu load) show no xruns here
even with relatively high cpu load (around 70%) and a periodsize of 64
frames. Of course i had the mandatory additional kernel compile running
in the background :)

flo

Oh wow. Just before hitting send i got three xruns of around
0.020-0.050msec. Ok, will read up on recent emails to see what to do to
debug these.